SCRS Conference Proceedings on Intelligent Systems
Performance Analysis of Preconditioner Based Image Reconstruction in High Resolution Microwave Tomography
Authors: N. Nithya and MSK. Manikandan
Publishing Date: 26-04-2022
ISBN: 978-93-91842-08-6
Abstract
The accuracy and speed of convergence in Microwave Tomography Imaging System (MwTIS) affected by ill-condition nature of coefficient matrix. Krylov subspaces based regularization methods are effectively produce the solution with ill-condition problems.The proposed work, details the preconditioner based regularization method called Enriched Conjugate Gradient Least Square (ECGLS) is good at handling ill-condition problem in MwTIS. In this paper, first, it analyzes the impact of frequency changes in the condition number of the coefficient matrix and evaluates the efficacy of ECGLS method.It achieves 37.42 % of MSE at 6 reconstruction iteration for brain dataset.
Keywords
Microwave Tomography, ill-condition problem, ECGLS, Brain Imaging, Regularization.
Cite as
N. Nithya and MSK. Manikandan, "Performance Analysis of Preconditioner Based Image Reconstruction in High Resolution Microwave Tomography", In: Raju Pal and Praveen Kumar Shukla (eds), SCRS Conference Proceedings on Intelligent Systems, SCRS, India, 2022, pp. 467-474. https://doi.org/10.52458/978-93-91842-08-6-44